I am getting response 429 for my deid service requests

Thanks for the response.

There are times where the number of requests exceed 100 in 5 seconds. So we would like to increase the capacity

I am not able raise support request. I get the message that "Your subscription has an internal support plan through Microsoft Azure. And I can't proceed further. I don't know how to get support using internal support plan.

I am Microsoft FTE and I am working for a project within Microsoft.

Thanks

Srivatsa

Hi @Srivatsa Ramakrishna ,

A 429 almost always means the service is limiting request throughput because you're sending more requests than allowed in a given time window.

1. Confirm it’s throttling (rate limiting)
   • 429 responses typically mean the service hit a defined rate limit and is temporarily rejecting further requests.
2. Analyze your request patterns (what’s driving the spike)
   • Check in Azure Monitor/metrics and/or diagnostic logs for when 429s happen and which operations trigger them.
   • Look for bursty patterns (many requests close together), “hot” workload areas (traffic concentrated on one partition/resource), and excessive metadata/management-style calls.
3. Reduce the effective request rate
   • If your client/app is making many requests in a short period, throttle yourself: add queueing/batching, reduce concurrency, and avoid redundant calls.
   • If your workflow includes retries, ensure you’re not creating a retry storm (rapid repeated retries worsen throttling).
4. Implement proper retry behavior
   • Use server guidance such as Retry-After (when present).
   • Add exponential backoff so retries wait longer each time instead of immediately hammering the service again.
5. Increase capacity / quotas if applicable
   • Review service limits/quotas in the Azure portal to see if you’re saturating allocated throughput.
   • If sustained traffic requires more, submit a quota/limit increase request.
   • Also verify whether you should spread workload across multiple resources (for some services, distributing across subscriptions/regions helps when the issue is localized).
6. If you’re on Azure Digital Twins (DEID) specifically
   • The recommended resource to review is the Azure Digital Twins service limits, which is where capacity/rate limits are defined.

Hope this helps!

If the resolution was helpful, kindly take a moment to click on and click on Yes for was this answer helpful. And, if you have any further query do let us know

Hello Srivatsa Ramakrishna,

Greetings! Thanks for raising this question in the Q&A forum.

The 429 responses you are seeing from the de-identification service are expected throttling behavior, not an error condition or misconfiguration. The de-identification service enforces a fixed rate limit at the data plane: requests are throttled if you exceed 1 MB per 5 seconds or 100 requests per 5 seconds. This limit applies per de-identification service instance and is not currently adjustable from the Azure Portal, which is why you are hitting it under sustained load.

Here is how to work through it:

Confirm you are actually hitting the documented threshold Check your request pattern against the two limits: total payload size per 5-second window and total request count per 5-second window. Requests can't exceed 50 KB, jobs can process no more than 10,000 documents, and each document processed by a job can't exceed 2 MB. If your workload sends many small synchronous requests rather than batching into a job, you will hit the request-count limit well before any data-size limit.

Switch high-volume workloads to the asynchronous job API where possible If you are calling the synchronous de-identify endpoint per document, batching documents into a single job submission (tag, redact, or surrogate) from Blob Storage reduces the number of individual requests counted against the 5-second window, since the job runs as one submission rather than hundreds of discrete calls.

Implement retry with backoff honoring the throttling response On a 429, back off and retry rather than immediately resubmitting, since immediate retries only add more requests to the same 5-second window and prolong the throttling.

Request a capacity increase through Azure Support Since this limit is fixed at the service level and not self-service adjustable, if your use case requires higher throughput, you need to submit a support request for consideration. Go to Help + support > New support request, set Issue type to Technical, Service to Azure Health Data Services, and in the problem description specify:

• The exact throughput you need (MB per 5 seconds and/or requests per 5 seconds)
  • Your current de-identification service resource name and region
    • A brief description of your workload pattern (synchronous calls vs job-based)

Since your support plan wasn't specified, note that a Technical ticket requires at least Developer-tier paid support; if you are on Basic support only, you can still open a Subscription Management request referencing this same capacity need, which routes to the product team for review at no cost.

If this answer helps you kindly accept the answer which will help others who have similar questions.

Best Regards,

Jerald Felix.